In [1]:
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
import scipy.stats as stats
import re
In [2]:
fatal_killings_raw = pd.read_csv("fatal_encounters_dot_org.csv",skiprows = [28622])
fatal_killings_raw.tail()
Out[2]:
Unique ID Subject's name Subject's age Subject's gender Subject's race Subject's race with imputations Imputation probability URL of image of deceased Date of injury resulting in death (month/day/year) Location of injury (address) ... A brief description of the circumstances surrounding the death Dispositions/Exclusions INTERNAL USE, NOT FOR ANALYSIS Intentional Use of Force (Developing) Link to news article or photo of official document Symptoms of mental illness? INTERNAL USE, NOT FOR ANALYSIS Video Date&Description Unique ID formula Unique identifier (redundant) Date (Year)
28616 28618 Santos Anthony Villegas 36 Male Hispanic/Latino Hispanic/Latino Not imputed https://fatalencounters.org/wp-content/uploads... 8/17/2020 700 Orange Street ... Santos Anthony Villegas was being sought by po... Pending investigation Intentional use of deadly force https://www.sbsun.com/2020/08/18/man-who-polic... No NaN 8/17/2020: Santos Anthony Villegas was being s... NaN 28618 2020
28617 28619 Maksim Aleksandrovich Krechetov 44 Male European-American/White European-American/White Not imputed NaN 8/17/2020 600 block NE. 120th Avenue ... Maksim Aleksandrovich Krechetov, 44, allegedly... Suicide Suicide https://www.startribune.com/after-standoff-aut... No NaN 8/17/2020: Maksim Aleksandrovich Krechetov, 44... NaN 28619 2020
28618 28620 Yulia Khort 47 Female European-American/White European-American/White Not imputed https://fatalencounters.org/wp-content/uploads... 8/17/2020 600 block NE. 120th Avenue ... Maksim Aleksandrovich Krechetov, 44, allegedly... Murder/suicide No https://www.startribune.com/after-standoff-aut... No NaN 8/17/2020: Maksim Aleksandrovich Krechetov, 44... NaN 28620 2020
28619 28616 Reuben 14 Male African-American/Black African-American/Black Not imputed NaN 8/18/2020 900 block Patterson Avenue ... Deputies were following the car down Patterson... Pending investigation Pursuit https://www.wfmynews2.com/article/news/local/w... No NaN 8/18/2020: Deputies were following the car dow... NaN 28616 2020
28620 28621 Danny Ray Spence 33 Male European-American/White European-American/White Not imputed https://fatalencounters.org/wp-content/uploads... 8/18/2020 Spring Creek Road ... Police were seeking Danny Spence for charges t... Pending investigation No https://www.wsmv.com/news/hickman_county/autho... No NaN 8/18/2020: Police were seeking Danny Spence fo... NaN 28621 2020

5 rows × 29 columns

In [3]:
#cleaning the age column

def clean_ages(age):
    # Search for opening bracket in the name followed by
    # any characters repeated any number of times
    if re.search("months$", str(age)):
        return round(float(age[:2].rstrip(" "))/12, 2)
    elif re.search("s$", str(age)):
        return float(age[:2].rstrip(" "))
    elif re.search("^[0-9]{2}-", str(age)):
        return round((float(age[:2]) + float(age[3:6]))/2, 2)
    elif re.search("^[0-9]{2}/", str(age)):
        return round((float(age[:2]) + float(age[3:6]))/2, 2)
    else:
        return float(age)
          
# Updated the city columns
fatal_killings_raw["Subject's age"] = fatal_killings_raw["Subject's age"].apply(clean_ages)
  
# Print the updated dataframe
fatal_killings_raw["Subject's age"].unique()
Out[3]:
array([ 17.  ,  21.  ,  19.  ,  23.  ,  24.  ,  53.  ,  30.  ,  29.  ,
        55.  ,  42.  ,  41.  ,  45.  ,  20.  ,  25.  ,  22.  ,  31.  ,
        36.  ,  28.  ,  35.  ,    nan,  26.  ,  68.  ,  16.  ,  49.  ,
        27.  ,  50.  ,  44.  ,  43.  ,  38.  ,  34.  ,  32.  ,  14.  ,
        15.  ,  18.  ,  33.  ,  57.  ,   1.  ,  88.  ,  37.  ,  40.  ,
        85.  ,  48.  ,  56.  ,  52.  ,  46.  ,  63.  ,  74.  ,  60.  ,
        59.  ,  51.  ,  69.  ,  10.  ,  13.  ,  39.  ,  47.  ,  66.  ,
        79.  ,  12.  ,  11.  ,  54.  ,  65.  ,  75.  ,   6.  ,   7.  ,
         5.  ,   3.  ,  72.  ,  58.  ,  71.  ,  80.  ,  78.  ,  61.  ,
        73.  ,  67.  ,  70.  ,  77.  ,  76.  ,   8.  ,   9.  ,  64.  ,
        62.  ,   4.  ,  83.  ,   2.  ,  89.  ,  21.5 ,   1.5 ,  49.5 ,
         0.25,  84.  ,  90.  ,  81.  ,  87.  ,   0.5 ,   0.75,   0.83,
        86.  ,  92.  ,   0.17,   0.58,  82.  ,   0.67,  91.  ,  95.  ,
       101.  , 107.  ,  97.  ,  93.  ,   0.33])
In [4]:
#Fill the nulls with median of the age
fatal_killings_raw["Subject's age"].fillna(fatal_killings_raw["Subject's age"].median(), inplace = True)
In [5]:
#Question 1.1
#How does the use of force vary across age groups? - Preparing the age bins
#Create a copy of the raw dataset
fatal_killings_age_subset = fatal_killings_raw.copy()

#Rename appropriate columns in the subset
fatal_killings_age_subset.columns = fatal_killings_age_subset.columns.str.replace("'","") 
fatal_killings_age_subset.rename(columns = {"Subjects age" : "age",
                                            "Cause of death" : "cause_death",
                                            "Subjects race" : "race",
                                            "Subjects race with imputations":"race_imp",
                                            "Date (Year)":"year",
                                            "Location of death (city)":"city",
                                            "Agency responsible for death" : "police_dept",
                                            "Symptoms of mental illness? INTERNAL USE, NOT FOR ANALYSIS":"mental_illness_symptoms"},inplace=True)

fatal_killings_age_subset["age_bin"] = pd.cut(fatal_killings_age_subset.age, bins = [0,10,18,35,45,64,80,110],labels=["Children",
                                                                                                                      "Teenagers",
                                                                                                                      "Early_adults",
                                                                                                                      "Early_middle",
                                                                                                                      "Late_middle",
                                                                                                                      "Senior",
                                                                                                                      "Super_senior"])
In [6]:
#Question 1.1
#How does the use of force vary across age groups? - Plotting the overall scenario

#Get the counts in a seperate dataframe
fatal_age_plot_data = fatal_killings_age_subset[["age_bin"]].value_counts().reset_index()
fatal_age_plot_data.rename(columns={0:"counts"},inplace=True)
fatal_age_plot_data.sort_values(by="age_bin",inplace=True)
fatal_age_plot_data

sns.set(rc={'figure.figsize':(9,5)})

age_plot = sns.barplot(data = fatal_age_plot_data,x = "age_bin", y = "counts").set(title = "Use of intentional force across age groups (2000 - 2020)",
                                                                                   xlabel = "Population age groups",
                                                                                   ylabel = "Instances of intentional force")
ax = plt.gca()

# Iterate through the subplot and set the value of individual bars
for p in ax.patches:
    ax.text(p.get_x() + p.get_width()/2., p.get_height(), '%d' % int(p.get_height()), 
            fontsize=12, color='red', ha='center', va='bottom')

# Set x label position

ax.xaxis.set_label_coords(0.5, -0.15)
In [7]:
#Question 1.1
#How does the use of force vary across age groups for European/White American people?
fatal_age_cau = fatal_killings_age_subset[fatal_killings_age_subset["race"] == "European-American/White"]
fatal_age_plot_cau = fatal_age_cau[["age_bin"]].value_counts().reset_index()
fatal_age_plot_cau.rename(columns={0:"counts"},inplace=True)
fatal_age_plot_cau.sort_values(by="age_bin",inplace=True)

#Plot data for European / American people
sns.set(rc={'figure.figsize':(9,5)})

age_plot = sns.barplot(data = fatal_age_plot_cau,x = "age_bin", y = "counts").set(title = "Use of intentional force across age groups (2000 - 2020) - European - Americans / White",
                                                                                   xlabel = "Population age groups",
                                                                                   ylabel = "Instances of intentional force")
ax = plt.gca()

#Iterate through the subplot and set the value of individual bars
for p in ax.patches:
    ax.text(p.get_x() + p.get_width()/2., p.get_height(), '%d' % int(p.get_height()), 
            fontsize=12, color='red', ha='center', va='bottom')

#Set x label position

ax.xaxis.set_label_coords(0.5, -0.15)
In [8]:
#Question 1.1
#How does the use of force vary across age groups for Black/African American people?
fatal_age_afr = fatal_killings_age_subset[fatal_killings_age_subset["race"] == "African-American/Black"]
fatal_age_plot_afr = fatal_age_afr[["age_bin"]].value_counts().reset_index()
fatal_age_plot_afr.rename(columns={0:"counts"},inplace=True)
fatal_age_plot_afr.sort_values(by="age_bin",inplace=True)

#Plot data for European / American people
sns.set(rc={'figure.figsize':(9,5)})

age_plot = sns.barplot(data = fatal_age_plot_afr,x = "age_bin", y = "counts").set(title = "Use of intentional force across age groups (2000 - 2020) - Black / African Americans",
                                                                                   xlabel = "Population age groups",
                                                                                   ylabel = "Instances of intentional force")
ax = plt.gca()

#Iterate through the subplot and set the value of individual bars
for p in ax.patches:
    ax.text(p.get_x() + p.get_width()/2., p.get_height(), '%d' % int(p.get_height()), 
            fontsize=12, color='red', ha='center', va='bottom')

#Set x label position

ax.xaxis.set_label_coords(0.5, -0.15)
In [9]:
#Question 1.1
#How does the use of force vary across age groups for Black/African American people?
fatal_age_his = fatal_killings_age_subset[fatal_killings_age_subset["race"] == "Hispanic/Latino"]
fatal_age_plot_his = fatal_age_his[["age_bin"]].value_counts().reset_index()
fatal_age_plot_his.rename(columns={0:"counts"},inplace=True)
fatal_age_plot_his.sort_values(by="age_bin",inplace=True)

#Plot data for European / American people
sns.set(rc={'figure.figsize':(9,5)})

age_plot = sns.barplot(data = fatal_age_plot_his,x = "age_bin", y = "counts").set(title = "Use of intentional force across age groups (2000 - 2020) - Hispanic / Latinos",
                                                                                   xlabel = "Population age groups",
                                                                                   ylabel = "Instances of intentional force")
ax = plt.gca()

#Iterate through the subplot and set the value of individual bars
for p in ax.patches:
    ax.text(p.get_x() + p.get_width()/2., p.get_height(), '%d' % int(p.get_height()), 
            fontsize=12, color='red', ha='center', va='bottom')

#Set x label position

ax.xaxis.set_label_coords(0.5, -0.15)
In [10]:
fatal_age_plot_ove = fatal_killings_age_subset[["year","race"]].value_counts().reset_index()
fatal_age_plot_ove = fatal_age_plot_ove[fatal_age_plot_ove["race"].isin(["European-American/White", 
                                                                         "African-American/Black", 
                                                                         "Hispanic/Latino"])]
fatal_age_plot_ove.sort_values(by = ["year","race"])
fatal_age_plot_ove.rename(columns = {0:"counts"},inplace = True)

fatal_age_line_plot_race = fatal_age_plot_ove.pivot(index = "year", columns="race",values = "counts")

fatal_age_line_plot_overall = fatal_killings_age_subset[["year"]].value_counts().reset_index()
fatal_age_line_plot_overall.rename(columns = {0:"Overall"},inplace = True)

fatal_age_line_plot_race = fatal_age_line_plot_race.merge(fatal_age_line_plot_overall,on="year")

#Plotting line plots for races
fatal_age_line_plot_race["Overall"].plot(label = "Overall")
fatal_age_line_plot_race["European-American/White"].plot(label = "European-American/White")
fatal_age_line_plot_race["African-American/Black"].plot(label = "African-American/Black")
fatal_age_line_plot_race["Hispanic/Latino"].plot(label = "Hispanic/Latino")
plt.legend(loc="upper left")
plt.xlabel("Year")
plt.ylabel("Count of fatal incidents")
plt.title("Fatal encounters across races over the years")
plt.xticks(range(0,len(fatal_age_line_plot_race.year)),fatal_age_line_plot_race.year, rotation = 60)
plt.show()
In [11]:
#How does the use of force vary across years for European / White Americans?
fatal_age_cau = fatal_killings_age_subset[fatal_killings_age_subset["race"] == "European-American/White"]
print(fatal_age_cau.columns)
fatal_age_plot_cau = fatal_age_cau[["year"]].value_counts().reset_index()
fatal_age_plot_cau.rename(columns={0:"counts"},inplace=True)
fatal_age_plot_cau.sort_values(by="year",inplace=True)

#Plot data for European / American people
sns.set(rc={'figure.figsize':(15,5)})

age_plot = sns.lineplot(data = fatal_age_plot_cau,x = "year", y = "counts")

age_plot.set(title = "Use of intentional force across age groups (2000 - 2020) - European-American/White",
             xlabel = "year",
             ylabel = "Instances of intentional force")

ax = plt.gca()
age_plot.set_xticks(fatal_age_plot_cau.year)
plt.show()

Index(['Unique ID', 'Subjects name', 'age', 'Subjects gender', 'race',
       'race_imp', 'Imputation probability', 'URL of image of deceased',
       'Date of injury resulting in death (month/day/year)',
       'Location of injury (address)', 'city', 'Location of death (state)',
       'Location of death (zip code)', 'Location of death (county)',
       'Full Address', 'Latitude', 'Longitude', 'police_dept', 'cause_death',
       'A brief description of the circumstances surrounding the death',
       'Dispositions/Exclusions INTERNAL USE, NOT FOR ANALYSIS',
       'Intentional Use of Force (Developing)',
       'Link to news article or photo of official document',
       'mental_illness_symptoms', 'Video', 'Date&Description',
       'Unique ID formula', 'Unique identifier (redundant)', 'year',
       'age_bin'],
      dtype='object')
In [12]:
#How does the use of force vary across years for Blacks/African Americans?
fatal_age_afr = fatal_killings_age_subset[fatal_killings_age_subset["race"] == "African-American/Black"]
fatal_age_plot_afr = fatal_age_afr[["year"]].value_counts().reset_index()
fatal_age_plot_afr.rename(columns={0:"counts"},inplace=True)
fatal_age_plot_afr.sort_values(by="year",inplace=True)

#Plot data for European / American people
sns.set(rc={'figure.figsize':(15,5)})

age_plot = sns.lineplot(data = fatal_age_plot_afr,x = "year", y = "counts")

age_plot.set(title = "Use of intentional force across age groups (2000 - 2020) - African-American/Black",
             xlabel = "year",
             ylabel = "Instances of intentional force")

ax = plt.gca()
age_plot.set_xticks(fatal_age_plot_afr.year)
plt.show()
In [13]:
#How does the use of force vary across years for Hispanic/Latinos?
fatal_age_his = fatal_killings_age_subset[fatal_killings_age_subset["race"] == "Hispanic/Latino"]
fatal_age_plot_his = fatal_age_his[["year"]].value_counts().reset_index()
fatal_age_plot_his.rename(columns={0:"counts"},inplace=True)
fatal_age_plot_his.sort_values(by="year",inplace=True)

#Plot data for European / American people
sns.set(rc={'figure.figsize':(15,5)})

age_plot = sns.lineplot(data = fatal_age_plot_his,x = "year", y = "counts")

age_plot.set(title = "Use of intentional force across age groups (2000 - 2020) - Hispanic / Latinos",
             xlabel = "year",
             ylabel = "Instances of intentional force")

ax = plt.gca()
age_plot.set_xticks(fatal_age_plot_his.year)
plt.show()
In [14]:
#Question1.4 - Which cities are at the helm of most killings?

fatal_city_plot_data = fatal_killings_age_subset.copy()
fatal_city_plot_data = fatal_city_plot_data[["city"]].value_counts().reset_index()
fatal_city_plot_data.rename(columns={0:"counts"},inplace=True)
fatal_city_plot_data.sort_values(by="counts",ascending = False,inplace=True)
fatal_city_plot_data = fatal_city_plot_data[0:10]
fatal_city_plot_data.sort_values(by="counts",ascending = True,inplace=True)

fig, ax = plt.subplots()

plt.hlines(y = fatal_city_plot_data.city,xmin = 0, xmax = fatal_city_plot_data.counts)
plt.scatter(fatal_city_plot_data.counts,range(0,len(fatal_city_plot_data.city)))
plt.title("Cities that have seen the most fatal incidents from 2000 till 2020")
plt.xlabel("Number of fatal incidents")
plt.ylabel("Cities")
ax.xaxis.set_label_coords(0.5, -0.15)
ax.yaxis.set_label_coords(-0.13, 0.5)
plt.show()
In [15]:
#Question1.4.1 - Which cities are at the helm of most killings for European American / Whites?

fatal_city_plot_data = fatal_killings_age_subset[fatal_killings_age_subset["race"]=="European-American/White"]
fatal_city_plot_data = fatal_city_plot_data[["city"]].value_counts().reset_index()
fatal_city_plot_data.rename(columns={0:"counts"},inplace=True)
fatal_city_plot_data.sort_values(by="counts",ascending = False,inplace=True)
fatal_city_plot_data = fatal_city_plot_data[0:10]
fatal_city_plot_data.sort_values(by="counts",ascending = True,inplace=True)

fig, ax = plt.subplots()

plt.hlines(y = fatal_city_plot_data.city,xmin = 0, xmax = fatal_city_plot_data.counts)
plt.scatter(fatal_city_plot_data.counts,range(0,len(fatal_city_plot_data.city)))
plt.title("Cities that have seen the most fatal incidents from 2000 till 2020 - European-American/White")
plt.xlabel("Number of fatal incidents")
plt.ylabel("Cities")
ax.xaxis.set_label_coords(0.5, -0.15)
ax.yaxis.set_label_coords(-0.13, 0.5)
plt.show()
In [16]:
#Question1.4.2 - Which cities are at the helm of most killings for African-American/Black?

fatal_city_plot_data = fatal_killings_age_subset[fatal_killings_age_subset["race"]=="African-American/Black"]
fatal_city_plot_data = fatal_city_plot_data[["city"]].value_counts().reset_index()
fatal_city_plot_data.rename(columns={0:"counts"},inplace=True)
fatal_city_plot_data.sort_values(by="counts",ascending = False,inplace=True)
fatal_city_plot_data = fatal_city_plot_data[0:10]
fatal_city_plot_data.sort_values(by="counts",ascending = True,inplace=True)

fig, ax = plt.subplots()

plt.hlines(y = fatal_city_plot_data.city,xmin = 0, xmax = fatal_city_plot_data.counts)
plt.scatter(fatal_city_plot_data.counts,range(0,len(fatal_city_plot_data.city)))
plt.title("Cities that have seen the most fatal deaths from 2000 till 2020 - African-American/Black")
plt.xlabel("Number of fatal incidents")
plt.ylabel("Cities")
ax.xaxis.set_label_coords(0.5, -0.15)
ax.yaxis.set_label_coords(-0.13, 0.5)
plt.show()
In [17]:
#Question1.4.3 - Which cities are at the helm of most killings for European American / Whites?

fatal_city_plot_data = fatal_killings_age_subset[fatal_killings_age_subset["race"]=="Hispanic/Latino"]
fatal_city_plot_data = fatal_city_plot_data[["city"]].value_counts().reset_index()
fatal_city_plot_data.rename(columns={0:"counts"},inplace=True)
fatal_city_plot_data.sort_values(by="counts",ascending = False,inplace=True)
fatal_city_plot_data = fatal_city_plot_data[0:10]
fatal_city_plot_data.sort_values(by="counts",ascending = True,inplace=True)

fig, ax = plt.subplots()

plt.hlines(y = fatal_city_plot_data.city,xmin = 0, xmax = fatal_city_plot_data.counts)
plt.scatter(fatal_city_plot_data.counts,range(0,len(fatal_city_plot_data.city)))
plt.title("Cities that have seen the most gunshot deaths from 2000 till 2020 - Hispanic/Latino")
plt.xlabel("Number of fatal incidents")
plt.ylabel("Cities")
ax.xaxis.set_label_coords(0.5, -0.15)
ax.yaxis.set_label_coords(-0.13, 0.5)
plt.show()
In [18]:
#Question1.5 - Which police departments are at the helm of most killings?

fatal_pd_plot_data = fatal_killings_age_subset.copy()
fatal_pd_plot_data = fatal_pd_plot_data[["police_dept"]].value_counts().reset_index()
fatal_pd_plot_data.rename(columns={0:"counts"},inplace=True)
fatal_pd_plot_data.sort_values(by="counts",ascending = False,inplace=True)
fatal_pd_plot_data = fatal_pd_plot_data[0:10]
fatal_pd_plot_data.sort_values(by="counts",ascending = True,inplace=True)
fatal_pd_plot_data

fig, ax = plt.subplots()

plt.hlines(y = fatal_pd_plot_data.police_dept,xmin = 0, xmax = fatal_pd_plot_data.counts)
plt.scatter(fatal_pd_plot_data.counts,range(0,len(fatal_pd_plot_data.police_dept)))
plt.title("Police departments that oversaw most fatal incidensts from 2000 till 2020")
plt.xlabel("Number of fatal incidents")
plt.ylabel("Police departments")
ax.xaxis.set_label_coords(0.5, -0.15)
ax.yaxis.set_label_coords(-0.30, 0.5)
plt.show()
In [19]:
#Question1.5.1 - Which police departments are at the helm of most killings of European-American / Whites?

fatal_pd_plot_data = fatal_killings_age_subset[fatal_killings_age_subset["race"]=="European-American/White"]
fatal_pd_plot_data = fatal_pd_plot_data[["police_dept"]].value_counts().reset_index()
fatal_pd_plot_data.rename(columns={0:"counts"},inplace=True)
fatal_pd_plot_data.sort_values(by="counts",ascending = False,inplace=True)
fatal_pd_plot_data = fatal_pd_plot_data[0:10]
fatal_pd_plot_data.sort_values(by="counts",ascending = True,inplace=True)
fatal_pd_plot_data

fig, ax = plt.subplots()

plt.hlines(y = fatal_pd_plot_data.police_dept,xmin = 0, xmax = fatal_pd_plot_data.counts)
plt.scatter(fatal_pd_plot_data.counts,range(0,len(fatal_pd_plot_data.police_dept)))
plt.title("Police departments that oversaw most fatal incidents involving European-American/Whites from 2000 till 2020")
plt.xlabel("Number of fatal incidents")
plt.ylabel("Police departments")
ax.xaxis.set_label_coords(0.5, -0.15)
ax.yaxis.set_label_coords(-0.30, 0.5)
plt.show()
In [20]:
#Question1.5.2 - Which police departments are at the helm of most killings of European-American / Whites?
fatal_pd_plot_data = fatal_killings_age_subset[fatal_killings_age_subset["race"]=="African-American/Black"]
fatal_pd_plot_data = fatal_pd_plot_data[["police_dept"]].value_counts().reset_index()
fatal_pd_plot_data.rename(columns={0:"counts"},inplace=True)
fatal_pd_plot_data.sort_values(by="counts",ascending = False,inplace=True)
fatal_pd_plot_data = fatal_pd_plot_data[0:10]
fatal_pd_plot_data.sort_values(by="counts",ascending = True,inplace=True)
fatal_pd_plot_data

fig, ax = plt.subplots()

plt.hlines(y = fatal_pd_plot_data.police_dept,xmin = 0, xmax = fatal_pd_plot_data.counts)
plt.scatter(fatal_pd_plot_data.counts,range(0,len(fatal_pd_plot_data.police_dept)))
plt.title("Police departments that oversaw most fatal incidents involving African-American/Blacks from 2000 till 2020")
plt.xlabel("Number of fatal incidents")
plt.ylabel("Police departments")
ax.xaxis.set_label_coords(0.5, -0.15)
ax.yaxis.set_label_coords(-0.30, 0.5)
plt.show()
In [21]:
#Question-2.2 - Did the victims have any mental illness symptoms / Drug and alcohol abuse?
fatal_killings_age_subset["mental_illness_symptoms"] = fatal_killings_age_subset["mental_illness_symptoms"].fillna("Unknown")
fatal_ms_plot_ove = fatal_killings_age_subset[["mental_illness_symptoms"]].value_counts().reset_index()
fatal_ms_plot_ove.rename(columns={0:"counts"},inplace=True)
fatal_ms_plot_ove.sort_values(by="counts",ascending = False,inplace=True)

fig, ax = plt.subplots(figsize=(9, 5))
sns.barplot(data=fatal_ms_plot_ove, x="counts", y="mental_illness_symptoms").set(title = "Mental illness symptoms / Drug and Alcohol use status of killed victims from 2000 till 2020",
                                                                                 xlabel = "Counts of Individuals",
                                                                                 ylabel = "Symptoms")
plt.show()
In [22]:
#Treemap for police dept involved in the most number of fatal deaths for African Americans
import plotly.express as px
fatal_pd_plot_data = fatal_killings_age_subset[fatal_killings_age_subset["race"]=="African-American/Black"].copy()
fatal_pd_plot_data = fatal_pd_plot_data[["police_dept"]].value_counts().reset_index()
fatal_pd_plot_data.rename(columns={0:"counts"},inplace=True)
fatal_pd_plot_data.sort_values(by="counts",ascending = False,inplace=True)
fatal_pd_plot_data = fatal_pd_plot_data[0:5]
fatal_pd_plot_data.sort_values(by="counts",ascending = True,inplace=True)
fatal_pd_plot_data
fig = px.treemap(fatal_pd_plot_data, path=['police_dept'],values='counts', width=800, height=400)
fig.update_layout(
    treemapcolorway = px.colors.sequential.Inferno, #defines the colors in the treemap
    margin = dict(t=50, l=25, r=25, b=25))
fig.show()
In [23]:
#Sunburst chart depicting the cities with the most fatal outcomes for Whites, Blacks and Latinos
fatal_city_plot_data = fatal_killings_age_subset.copy()
fatal_city_plot_data = fatal_city_plot_data[fatal_city_plot_data["race"].isin(["European-American/White",
                                                                              "African-American/Black",
                                                                              "Hispanic/Latino"])]
fatal_city_plot_data = fatal_city_plot_data[["race","city"]].value_counts().reset_index()
fatal_city_plot_data.rename(columns={0:"counts"},inplace=True)
fatal_city_plot_data.sort_values(by="counts",ascending = False,inplace=True)

fatal_city_plot_data_white = fatal_city_plot_data.loc[fatal_city_plot_data.race=="European-American/White"].copy()
fatal_city_plot_data_white.sort_values(by="counts",ascending = False,inplace=True)
fatal_city_plot_data_white = fatal_city_plot_data_white[0:3]

fatal_city_plot_data_black = fatal_city_plot_data.loc[fatal_city_plot_data.race=="African-American/Black"].copy()
fatal_city_plot_data_black.sort_values(by="counts",ascending = False,inplace=True)
fatal_city_plot_data_black = fatal_city_plot_data_black[0:3]

fatal_city_plot_data_latino = fatal_city_plot_data.loc[fatal_city_plot_data.race=="Hispanic/Latino"].copy()
fatal_city_plot_data_latino.sort_values(by="counts",ascending = False,inplace=True)
fatal_city_plot_data_latino = fatal_city_plot_data_latino[0:3]

final_city_data_plot = pd.concat([fatal_city_plot_data_white,fatal_city_plot_data_black,fatal_city_plot_data_latino])
print(final_city_data_plot)
fig = px.sunburst(
    final_city_data_plot,
    path = ['race','city'],
    values = 'counts',
    color = px.colors.sequential.r[0:9]
)

fig.show()

                       race         city  counts
11  European-American/White      Houston      76
13  European-American/White    Las Vegas      74
16  European-American/White      Phoenix      70
0    African-American/Black      Chicago     249
2    African-American/Black      Houston     170
5    African-American/Black  Los Angeles     107
1           Hispanic/Latino  Los Angeles     174
3           Hispanic/Latino  San Antonio     119
4           Hispanic/Latino      Houston     116
In [24]:
#Heatmap across time that captures killings across age groups
fatal_time_age_data = fatal_killings_age_subset.copy()
fatal_time_age_data = fatal_time_age_data.loc[fatal_time_age_data["race"] == "African-American/Black"].copy()
fatal_time_age_data = fatal_time_age_data[["year","age_bin"]].value_counts().reset_index()
fatal_time_age_data.rename(columns={0:"counts"},inplace=True)
fatal_time_age_data.sort_values(by = ["year","age_bin"],inplace = True)
fatal_time_age_data.counts = fatal_time_age_data.counts.apply(lambda x:x/100)

heatmap_df = fatal_time_age_data.pivot_table(values = "counts", index = "age_bin", columns="year")
heatmap_df
sns.heatmap(heatmap_df, annot=True)
plt.title("Fatal incidents(in Hundreds) across age bins over years",fontdict = {"fontsize": 18, "fontweight" : "bold"}, pad = 20)
plt.xlabel("Years",fontdict = {"fontsize": 14, "fontweight" : "bold"}, labelpad = 10)
plt.ylabel("Age bins",fontdict = {"fontsize": 14, "fontweight" : "bold"}, labelpad = 10)
plt.show()
In [ ]: